Event-by-event analysis?

ROOT
1

I’ve got some proton tracking data from planes of silicon strip detectors. The detectors are in X-Y pairs which can be treated as a single 2-D plane. I have 6 of these planes, and I want to be able to look at maps of hits on an event-by-event basis. I’ve tried 2-D TGraphs, and I’ve gotten close, unfortunately the problem is that the points on the plots are only about a pixel in size.

Ideally I was thinking to either superimpose the 6 planes in different colors, or have a pad split into 6 graphs (one for each plane), but I’ve become so frusterated that I’ve dumbed it down to this:

I now just want to use the graph’s horizontal axis to represent the plane and the vertical axis to represent the channel, thus looking at only one dimension at a time.

192-chan y-axis
| x…x
|
| …x
| …x
| x…x…x
| …x
| …x…x

. 1 . 2 . 3 . 4 . 5 . 6
6-plane x-axis

This way I would be able to visualize that planes 1 and 2 which are noisy, needed certain selections for their events, etc.

Anyone have any ideas on this?

Thank you!
-Dominic Lucia
SCIPP, Santa Cruz, CA

2

Can you provide:
[ul]a small data file
the script producing your best attempt[/ul]
so that we can tweak it appropriately.

Cheers,
Philippe

3

Thanks for the prompt reply, but I became distracted with another project.

This is something I whipped up, but its not really even close to what I want.

Something to remember is each hit constitutes its own line in the .root file database, but obviously there are more than one hits per event. This is why each hit in the same event has the same event_id (UInt) value.

Example files (no beam) are on my webspace at
scipp.ucsc.edu/~dominic/pCT/roothelp/

The smaller file will give you an idea of how our data is stored and the larger file actually has an appreciable amount of hits.

Thank you for your time!
-Dominic Lucia
SCIPP, Santa Cruz, CA
quickscript.cpp (1.02 KB)

4

Hi,

Your general structure seems okay for what you want to do.

Note that usually, the hits in an event are represented by using containers (or variable size arrays) so that there is one entry per event. This makes things easier.

[quote]This is something I whipped up, but its not really even close to what I want.[/quote]Somehow I am unable to extract from your comments what you actually want to see. I re-read your code and saw:

[quote] I’ve tried 2-D TGraphs, and I’ve gotten close, unfortunately the problem is that the points on the plots are only about a pixel in size. [/quote]You can very simply change the points size by calling SetMarkerSize.

Cheers,
Philippe.

5

[quote=“pcanal”]
Note that usually, the hits in an event are represented by using containers (or variable size arrays) so that there is one entry per event. This makes things easier.[/quote]
Is it possible to change the format of my root file? Could I write a script to extract the information and then reinput it in a more usable format?

Thanks!
-Dominic

6

Im now working on a script to change the format of our file.

What would you reccomend as far as organization? I was thinking something as simple as having a 6x128 array (6 modules by 128 channels) of Int_t with a 1 signifying a hit.

For the puposes of the analysis I’m doing, I think this might be the easiest and best way.